Technology Computer‐Aided Design for Power Semiconductor Device Simulation: Physical Models, Numerical Methods and Software Platforms
High Voltage · 2026-10 · 11 (5) · 1501-1523
DOI: 10.1049/hve2.70260 ↗Kurz gefasst
Fachartikel in der Zeitschrift High Voltage, online erschienen am 06.11.2026. Verfasst von Qingyuan Shi und 2 weiteren Autor:innen. Die Arbeit nennt 218 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Modern high‐voltage power systems increasingly rely on power semiconductor devices for energy conversion, control and transmission. Device design, optimisation and failure analysis require knowledge of the internal distributions of electric field, carrier density and temperature, which are difficult to obtain directly from terminal electrical measurements. Within technology computer‐aided design, device simulation calculates these quantities by solving coupled semiconductor equations for specified device structures and operating conditions. This paper reviews the physical models, numerical methods and software platforms used in power semiconductor device simulation. First, power semiconductor device models are classified into five levels, ranging from ideal switch representations for system‐level studies to models based on partial differential equations that calculate internal physical quantities. For device simulation based on partial differential equations, the review covers the governing electrothermal equations, semiconductor physical models and device–circuit coupling. Numerical methods are then organised into four stages: spatial discretisation, time integration, nonlinear iteration and the solution of sparse linear systems, with emphasis on discretisation accuracy, stability and conservation, as well as solver robustness, computational cost and parallel scalability. The review traces the evolution of device simulation software from early academic programs through industry‐driven expansion to market consolidation among a small number of commercial platforms and identifies a critical gap in the software landscape: few platforms combine realistic power device simulation capabilities with accessible and extensible implementations of numerical methods. Beyond this gap, emerging semiconductor materials, simulations at larger scales and severe electrothermal stresses create additional challenges for software development. Future device simulators will need to combine reliable physical models with robust and scalable solvers and progressively incorporate artificial intelligence into simulation and design.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Shi, Q., Zhuang, C., & Zeng, R. (2026). Technology Computer‐Aided Design for Power Semiconductor Device Simulation: Physical Models, Numerical Methods and Software Platforms. High Voltage, 11(5), 1501-1523. https://doi.org/10.1049/hve2.70260
Vancouver
Shi Q, Zhuang C, Zeng R. Technology Computer‐Aided Design for Power Semiconductor Device Simulation: Physical Models, Numerical Methods and Software Platforms. High Voltage. 2026;11(5):1501-1523. doi:10.1049/hve2.70260
BibTeX
@article{Shi2026,
author = {Shi, Qingyuan and Zhuang, Chijie and Zeng, Rong},
title = {Technology Computer‐Aided Design for Power Semiconductor Device Simulation: Physical Models, Numerical Methods and Software Platforms},
journal = {High Voltage},
year = {2026},
volume = {11},
number = {5},
pages = {1501--1523},
doi = {10.1049/hve2.70260}
}
EUVIMED